The approaches described in this section are approaches that could be pursued, but not necessarily approaches that have been previously conceived or pursued. Therefore, unless otherwise indicated, it should not be assumed that any of the approaches described in this section qualify as prior art merely by virtue of their inclusion in this section.
Existing biometric identification and authentication methods usually involve verification of values of a single parameter against some norms or baseline data. The norms or the baseline data may be determined based on the averaged values computed for a population of individuals exhibiting similar characteristics. However, as the requirements for the authentication and security systems become more rigorous, relying on the norms or baselines often becomes insufficient and inadequate.
As computer technologies become more and more sophisticated, new demands are placed on authentication and security systems. However, the conventional authentication and security systems are often unable to keep up with the new demands. These days, when even quite complex conventional security systems can be compromised, the conventional systems appear to be unreliable or obsolete. For example, some of the conventional security systems may be unable to distinguish the falsely identified breaches (also referred to as “false positive”) from the real security breaches.